In certain oil reservoirs, the pressure inside the reservoir is insufficient to push wellbore fluids to the surface without the help of a pump or other so-called artificial lift technology such as gas lift in the well. With a gas-based artificial lift system, external gas is injected into special gas lift valves placed inside a well at specific design depths. The injected gas mixes with produced fluids from the reservoir, and the injected gas decreases the pressure gradient inside the well, from the point of gas injection up to the surface. Bottom hole fluid pressure is thereby reduced, which increases the pressure drawdown (pressure difference between the reservoir and the bottom of the well) to increase the well fluid flow rate.
Other artificial lift technologies may also be used, e.g., centrifugal pumps such as electro-submersible pumps (ESPs) or progressing cavity pumps (PCPs). Furthermore, with some oil reservoirs, a mixture of artificial lift technologies may be used on different wells.
During the initial design of a gas lift or other artificial lift system to be installed in a borehole, software models have traditionally been used to determine the best configuration of artificial lift mechanisms, e.g., the gas lift valves, in a well, based on knowledge about the reservoir, well and reservoir fluids. However, models that are limited to single wells generally do not take into account the effects of other wells in the same field, and it has been found that the coupling through the surface network of wells in the same field will affect the actual rates experienced by each well.
Software models have also been developed to attempt to optimally configure artificial lift mechanisms for multiple wells coupled to each other in the same oilfield or surface production network. Such models, which may be referred to as surface network models, better account for the interrelationships between wells and the artificial lift mechanisms employed by the various wells. Nonetheless, shortcomings still exist with such multi-well models. For example, a surface network model is an approximation to reality, so the computed optimized lift gas rates for a gas-based artificial lift system are an approximation to the true optimum rates. In addition, a surface network model generally has to be continually re-calibrated so that it remains an accurate representation of the real network. Online measurements of a surface production network (e.g., actual measurements of pressures, temperatures and flow rates) generally are cross-checked against model calculations to insure that the two are consistent. If they differ substantially, a human operator may intervene to alter the surface network model to improve the match. In addition, in some instances a surface network model may have to be re-run whenever surface network conditions change, that is, whenever the well head flowing back pressures change, so that optimized lift gas rate values change. Surface network conditions can change frequently, for example, in response to instantaneous changes in the surface facility settings, equipment status and availability (equipment turning on and off), changes in ambient temperature, and at slower time scales, changes in fluid composition such as gas-oil ratio and water cut and surface network solid buildup or bottle-necking.
Moreover, another problem arising as a result of the use of surface network models is the need for centralized computation or determination of optimal artificial lift parameters for wells in a surface network. In many cases, set points for individual well gas lift flow rate values are calculated by a central controller and communicated to the individual wells, where closed loop well controllers maintain the desired gas lift flow rate set points, in the absence of any feedback or other operating conditions being experienced by the wells. As such, the centralized nature of the model calculations is not particularly responsive to the actual conditions for each well.
Therefore, a need continues to exist in the art for an improved manner of optimizing artificial lift technologies for multiple wells in a multi-well production network.